Dr. Hossein Beygi Nasrabadi, Prof. Harald Sack

Ontologies play a central role in materials science and engineering (MSE), enabling structured, machine-readable representation of materials knowledge across experiments, processes, and data platforms. A key challenge is interoperability: materials data is produced by heterogeneous sources and must be exchanged and integrated consistently. Reusing well-designed high-level ontologies supports interoperability, consistency, and long-term maintainability. High-level here refers to top- and mid-level ontologies, which define the most general concepts from which domain ontologies are derived. In MSE, the Elementary Multiperspective Material Ontology (EMMO) is a well-known top-level ontology, providing a philosophically grounded framework for representing materials, processes, and measurements. Following a more standardization-oriented approach that enables cross-domain interoperability and integration, the PMD core ontology (PMDco) has been developed as a mid-level ontology based on the ISO/IEC 21838-2:2021 standard. PMDco v3, is a highly advanced ontology and remains under active development. This master thesis investigates the matching between EMMO and PMDco ontologies. Using LLMs, it studies which classes match lexically and semantically. For matched classes, EMMO annotation comments will be collected and added to PMDco v3; for unmatched classes, the thesis will compile classes, definitions, and EMMO comments and suggest hierarchy extensions for PMDco.